Exploring the Abilities and Limitations of the SARIMA Model For Retail Demand Forecasting on Dillard’s Data From 2021-2025

Retail demand forecasting is important for businesses such as Dillard’s, a leading American department store chain, because it allows for data-driven decisions that enhances profitability, efficiency, and customer satisfaction. These forecasts serve as a strategic foundation for various business aspects from inventory management and supply chain planning to marketing campaigns and more. Forecasting methods vary widely from traditional to advanced machine learning models, each offering their pros and cons. This thesis will explore the ability and limitations of the seasonal autoregressive integrated (SARIMA) model to forecast demand on nearly four years of product group data provided by Dillard’s.